root @5.34.05: add 'xrootd' variant

[cjones051073 (Chris Jones)]

Hi,

This patch adds a new xrootd variant to ROOT. Disabled by default, when enabled uses a new xrootd port to build with xrootd support.

Relies on ticket #38341 and ticket #38342

Also adapts to the changes in the pythia include file location, in the above ticket, that resolves the conflict between pythia and libevent (which xrootd uses).

cheers Chris

[andre.david@…]

Cc Me!

[andre.david@…]

Took me a while to figure out that Portfiles are now under ${pfx}/var/macports/sources/rsync.macports.org/release/tarballs/ instead of ${pfx}/var/macports/sources/rsync.macports.org/release/ports (should I remove the latter?) but when I did, the patch could be applied without a hitch.

The xrootd dependence on libevent caused some trouble at the beginning:

--->  Installing libevent @2.0.21_0
--->  Activating libevent @2.0.21_0
Error: org.macports.activate for port libevent returned: Image error: /opt/local/include/event.h already exists and does not belong to a registered port.  Unable to activate port libevent. Use 'port -f activate libevent' to force the activation.
Error: Failed to install libevent
Please see the log file for port libevent for details:
    /opt/local/var/macports/logs/_opt_local_var_macports_sources_rsync.macports.org_release_tarballs_ports_devel_libevent/libevent/main.log

but eventually doing port -f activate libevent did the trick and the rest of the build proceeded uneventfully.

I confirm that xrootd and root are working as expected. Cheers!

[cjones051073 (Chris Jones)]

The problem with /opt/local/include/event.h is due to a conflict between the pythia and the libevent ports, that both currently install a file with that name.

Ticket #38341 is designed to fix this, by moving the pythia includes to /opt/local/include/pythia. Note that the patch submitted here for root assumes that pythia patch has also been applied, so you should also apply it, if you wish to use pythia with ROOT.

Anyway, thanks for the feedback and confirmation it works.

Chris

[andre.david@…]

BTW - while I run with the Pythia #38341 patch - what does #38342 have to do with all this?

[cjones051073 (Chris Jones)]

You would not have had the conflict error if you where using the patch in ticket #38341. Are you sure you fully rebuilt pythia and root with both patches ?

Sorry, looks like a misquoted the ticket numbers. This indeed has nothing to do with ticket #38342.

Just to be clear, this patch relies in ticket #38341 to update pythia, and ticket #38343 for the new xrootd port (now commited).

Chris

[andre.david@…]

I confirm that the patches in #38344 (this ticket) and #38341 correctly pick up and build the xrootd port from #38343, and all works as expected.

The reporter of #38236 is very happy with the outcome. ;-)

[cooljeanius (Eric Gallager)]

Cc Me!

[cooljeanius (Eric Gallager)]

Replying to andre.david@…:

The reporter of #38326 is very happy with the outcome. ;-)

Wait that's me... what does the objc port have to do with this? Or did you put the wrong issue number by accident?

[larryv (Lawrence Velázquez)]

Replying to egall@…:

Wait that's me... what does the objc port have to do with this? Or did you put the wrong issue number by accident?

Probably meant #38236.

[cjones051073 (Chris Jones)]

I mis quoted the ticket numbers in the original post. The correct ones are in comment # 5

[cjones051073 (Chris Jones)]

Replying to larryv@…:

Replying to egall@…:

Wait that's me... what does the objc port have to do with this? Or did you put the wrong issue number by accident?

Probably meant #38236.

Andre also mis quoted the ticket number here as well... He was indeed referring to ticket #38236

[larryv (Lawrence Velázquez)]

For future reference, all Trac fields take proper email addresses, not weird MacPorts-formatted ones.

[larryv (Lawrence Velázquez)]

Committed in r104095.

[andre.david@…]

Replying to jonesc@…:

Replying to larryv@…:

Replying to egall@…:

Wait that's me... what does the objc port have to do with this? Or did you put the wrong issue number by accident?

Probably meant #38236.

Andre also mis quoted the ticket number here as well... He was indeed referring to ticket #38236

Yes, yes. Sorry everyone. I just corrected the original comment.

[larryv (Lawrence Velázquez)]

Replying to andre.david@…:

Yes, yes. Sorry everyone. I just corrected the original comment.

But we’ll always know. YOU CAN’T HIDE FROM THE PAST